70 mm 5 6 7 8 1 2 3 4 Figure 1: Location of longitudinal tensile specimens in WM and HAZ Top Mid- section Bottom This was performed by direct measurements of the diameter reduction during testing. The measurements were performed in two orthogonal directions normal to the tensile axis using 4 displacement gauges in a special designed fixture. FINITE ELEMENT ANALYSES FE Analyses were carried out to investigate the stress-strain state and the void coalescence (ductile fracture). The modified Gurson model were implemented into the FE Program ABAQUS version 5.8. All the specimens were modeled with axisymmetric quadratic 8 node elements. A reduced integration scheme was adopted. Due to symmetry conditions, only one quarter of the specimens was modeled. The nodes at the top end of the model were used to prescribe a monotonic vertical displacement to simulate the uniaxial loading situation. The material models were defined by representative Bridgeman corrected true stress strain data from the smooth specimens, with modulus of elasticity E=210000 MPa and Poisson's ratio ν=0.3. Since no Bridgeman corrected stress strain curve could be obtained for the transversal direction the transversal fitting procedure was carried out using stress-strain curves from the longitudinal direction. RESULTS Base material Figure 2a shows and overview of load vs. diameter reduction for all base material specimens. Figure 2b shows simulated curves for notch radius R = 1.0 mm fitted by different values of the Gurson parameter f0. The results clearly show that increased stress triaxiality represented by decreasing notch radius causes a rise in the load level and a lowering of the diameter reduction at fracture. The fitted curves for R=1.0 mm agrees well with the experiment for f0 = 0.0001. After performing the fitting procedure for all geometries, the results were evaluated in ductility diagrams, comparing the measured and simulated strain at the beginning of final fracture, εc (Figure 3) In the experiments the critical strain was taken as the strain at maximum true stress. In the FE-analysis εc was represented by the strain at the initiation of failure, represented by a sudden load drop. Critical strain is plotted as a function of notch radius. Black points represent experimental values. Simulated results are represented by curves. The best fit for the base metal was obtained for the cluster nucleation model with f0 = 0.0001. The continuos model with A= 0.0005 also fitted the experimental results reasonably well, except for the smooth and sharpest notched specimens. It can be noticed that the model predicts higher ductility for the R0.4 than for the R1.0 specimens. The experimental results, however, showed relatively similar ductility values for these two geometries.
RkJQdWJsaXNoZXIy MjM0NDE=